Association between parents’ smoking status and tobacco exposure in school-age children: assessment using major urine biomarkers

Children are at risk of exposure to secondhand smoke. We aimed to evaluate the extent of their exposure to it in relation to their parents’ smoking status by using biomarkers relevant to smoking. We evaluated 847 school-age children (6–12 years) who lived with their parents, using data from the Korea National Health and Nutrition Examination Survey 2016–2018. Secondhand smoke exposure in children of non-smoking and smoking parents was assessed by measuring urinary 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol (NNAL) and cotinine concentrations. Overall, the parents of 482 (55.1%) children smoked and those of 392 (44.9%) children did not smoke. After adjusting for covariates, significantly higher concentrations of NNAL (β = 0.482, standard error [S.E.] = 0.065, P < 0.001) and cotinine (β = 0.472, S.E. = 0.06, P < 0.001) were found in children of smoking parents than in children of non-smoking parents. Children of parents who smoked a higher number of cigarettes showed higher NNAL and cotinine concentrations than children of non-smoking parents. Children with both parents who smoked showed the highest NNAL and cotinine concentrations. Children of smoking parents are at a higher risk of exposure to secondhand smoke. A smoke-free environment must be maintained to protect children from the harmful effects of secondhand smoke. Therefore, comprehensive national anti-smoking policies are required.

the house and spread indoors, causing SHS or THS exposure 14 . Therefore, children can still be exposed to toxic substances from smoking.
The major biomarkers of SHS and THS include 4-(methylnitrosoamino)-(3-pyridyl)1-butanol (NNAL) and cotinine, which are metabolites of nicotine 15 . Cotinine is one of the most commonly used tobacco exposure biomarkers in SHS and THS studies and has the advantage of very high specificity and sensitivity in screening tests 15,16 . NNAL is one of the metabolites of N-nitrosamines, a carcinogen derived from nicotine, and has the advantage of a long half-life of approximately 3 weeks 17 . Many studies have investigated the association between parents' smoking status, SHS and THS exposure biomarkers in children 11,14,15,18 . However, most previous studies have focused on adolescents in terms of smoking probability or on children with certain diseases 13,19 .
Thus, this study aimed to investigate the association between parental smoking and NNAL and cotinine concentrations, as biomarkers of SHS and THS exposure, in children. Furthermore, children's NNAL and cotinine concentrations were investigated according to the number of cigarettes smoked and smoking patterns in parents. We targeted school-age children living with their parents and have a low likelihood of smoking on their own, based on the premise that children who live with at least one smoking parent are more likely to be exposed to SHS and THS.

Discussion
Most studies on the association between parents' smoking status, SHS and THS exposure biomarkers have focused on adolescents in terms of smoking probability or on children with certain diseases. In this study of school-age children in Korea, at least one out of two children was living with a parent who smoked, and the NNAL and cotinine concentrations were higher in children whose parents were smokers. Further analysis confirmed that the higher the number of cigarettes smoked by both parents, the higher the degree of SHS exposure in children. To the best of our knowledge, no previous study in Asia has evaluated both NNAL and cotinine levels in school-age children to determine the extent of SHS exposure.
Our results are similar to those of previous studies 13,20 . However, it should be noted that our study targeted school-age children. School-age children are particularly vulnerable to second-hand smoke compared to other ages 11 . The adverse effects of parents smoking on their children's health, such as respiratory symptoms, can be reduced as their children grow older, due to spending less time with the parents 21 . Adolescents tend to spend more time outside the home, school-age children spend a lot of time at home and stay close to their parents, suggesting that living with parents who smoke can be a strong predictor of increased exposure to substances included in cigarettes. The highest NNAL and cotinine concentrations were observed in children when both parents smoked. These results were similar to those of previous studies 22,23 . Additionally, although a direct correlation is difficult, compared to when both the parents were non-smokers, the children's NNAL and cotinine concentrations were higher when only the mother smoked than when only the father smoked. Compared to children of non-smokers, children whose mothers alone smoked or whose both parents were smokers were 2-13 times more likely to be exposed to SHS at home 24 . However, a recent study showed that SHS exposure among adolescents is associated with paternal smoking 17 . In fact, worldwide smoking rates are higher for men than for women 25 . Thus, smoking abstinence by paternal figures is often chosen as the first strategy to reduce children's exposure to indirect smoking 13,26 . However, for school-age children, time spent with the mother tends to be more than two-fold longer than the time spent with the father. Therefore, smoking abstinence in mothers should also  24,28,29 , we found that the higher the number of cigarettes smoked, the higher the children's NNAL and cotinine concentrations, regardless of the child's age. This could be because as the number of cigarettes smoked by the parents increased, the amount of harmful substances adhering to their clothes and skin also rose, indirectly exposing the children.
In the subgroup analysis, the association trend was significant according to sex (male and female), BMI (underweight, normal weight, and overweight), household income (Q1, Q2, Q3, and Q4), and parents' education level (middle school or lower and college or higher). In the case of BMI, the NNAL and cotinine concentrations showed a tendency to be highest in the overweight group, but only the cotinine concentration was statistically significant. This is mainly due to the fact that NNAL is produced by smoking, whereas cotinine may be affected by diet 30 . We assumed that obese children consume more food than children of a normal weight, and that the amount of cotinine accumulated through food intake may influence the statistical significance. More detailed research on cotinine and food intake is needed in the future. NNAL and cotinine concentrations were both highest in children from the low-income group. People from this group have less awareness regarding the risks of exposure to SHS; thus, these children may be more vulnerable 31 . The children of parents with a higher level of education had lower NNAL and cotinine concentrations. This supports the premise that education level has a greater influence on SHS exposure than income, and individuals with higher education levels are less likely to smoke and, in cases where they do, are more likely to quit 31,32 .
Our study shows that even after controlling SHS exposure at home and in public places, many children are still likely to have high levels of NNAL, cotinine concentration due to their parents' smoking habits. This indicates that while the prohibition of smoking at home and in public show a highly negative correlation with children's exposure to SHS 29,31 , these policies alone cannot fully protect them from the adverse effects of SHS exposure due to parental smoking. This can be explained based on the results of previous studies in which children living with smoker parents had higher cotinine and NNAL concentration than children with non-smoker parents even if they do not smoke at home 15,33 . In other words, these results are the effect of THS, which allows children to inhale harmful substances by combining household fibers, clothing, sedimentation dust and surfaces with toxic substances related to external cigarettes, even if parents control their exposure at home 15 . Parents' smoking may cause substances such as tobacco-specific nitrosamine (TSNA) to be adsorbed by all indoor home surfaces, which then release these substances into the air 34,35 . Several studies have warned that THS is as damaging as SHS because it releases harmful substances similar to those released through SHS, causing DNA mutations and damage 34,36 .
The World Health Organization states that there is no safe level of exposure to SHS and THS a pollutant that causes serious illnesses in adults and children. Hence, the only effective way to protect the population from the harmful effects of exposure to SHS and THS is to maintain a 100% non-smoking environment 37 . Implementing physical measures or anti-smoking measures at home, such as the opening of windows or doors or removing cigarette smoke using a ventilator fan, is ineffective in preventing children's exposure to cigarette smoke. This is because only a completely non-smoking environment can prevent SHS and THS exposure in the home 11,33 . Parents' smoking cessation eliminates threats to their own and their children's health, so relevant policies should be encouraged. For this, it is first necessary to raise awareness of the risks of SHS and THS. It was found that the rate of smoking cessation attempts increased after the campaign to revitalize the hazard awareness of SHS and THS 38 . Even if the adverse effects of THS cannot be completely eliminated, these efforts can increase openness to laws prohibiting smoking in the home, and furthermore, the rate of successful cessation will be even higher if quit smoking intervention policies are implemented together 38,39 . Raising awareness of SHS and THS could be an effective strategy to protect children from tobacco exposure.
This study has some limitations. First, we used cross-sectional data. Therefore, the cause and effect and the direction of the relationships observed cannot be determined. Second, the results of this study were based on self-reported data. In this self-report, a vaper may report him or herself as a smoker despite not using flammable cigarettes, or a "social smoker" might report as a non-smoker. Thus, the number of cigarettes smoked may have been underestimated or overestimated, and some survey questions may be subject to recall bias. As a result, we cannot eliminate the likelihood that some smokers will be classified as nonsmokers or nonsmokers as smokers. Third, despite our efforts to control for confounding factors, not all covariates affecting NNAL and cotinine concentrations may have been considered. Lastly, This study sample derived from KNHANES was Table 1. General characteristics of the study population. BMI body mass index, IQR interquartile range, NNAL 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol, S.D. standard deviation, SHSE secondhand smoke exposure. *Age is a continuous variable and N = Mean, % = standard deviation. Despite these limitations, our study has important implications. This study evaluated the Association between parental smoking and children's exposure to SHS and THS using a well-defined nationally representative data in Korea. Our findings also support previous results. We targeted school-age children and thus minimized the bias related to smoking status. Further, we controlled for both SHS exposure in public and at home. These factors were not well-considered in previous studies. Furthermore, while analyses based on cotinine measurements   Our study demonstrated that children with parents who smoked are at a higher risk of exposure to SHS and THS, implying that individual efforts to avoid smoking in the presence of children may be an insufficient alternative. The best way to protect children from toxic substances from exposure to smoking is to quit smoking. This requires comprehensive anti-smoking arbitration policies, such as improving awareness of how to protect children from smoking substances.

Methods
Study population. This study was based on data from the 2016-2018 KNHANES VII and the secondary analysis of a large dataset. The KNHANES is a nationwide population-based cross-sectional survey conducted annually since 1998 under the direction of the Centers for Disease Control and Prevention of the Ministry of Health and Welfare to accurately assess the national health and nutritional status 40 .
The total number of respondents for the 2016-2018 KNHANES was 24,269. Participants included single parents, but if they did not match the criteria for a parent-child relationship (n = 1,384); had no data on age (n = 21,019) for those between 6 and 12 years; and those without data on NNAL levels, cotinine levels, or other independent variables (n = 992) were excluded. Finally, a total of 847 participants were included in the study (Fig. 2). KNHANES data is publicly accessible and ethical approval is not required for the use of the data. This data were collected with prior consent before participating in the survey and respondents' information was completely anonymized for use for research purposes 40 . . Adjusted for children's sex, age, body mass index, secondhand smoke exposure (house), and secondhand smoke exposure (public), and for parents' household income, type of housing, region, age, education level, private health insurance, drinking status, and year of evaluation. The reference group is the group of parents who are nonsmokers.   17,20 . The limit of detection (LOD) was 0.27399 ng/mL for cotinine and 0.1006 pg/mL for NNAL, and was calculated by dividing it by 2 referring to prior study 41,42 . In addition, a random subsample (1/2 or 1/3 of the total sample) of subjects aged 6 years or older was used for gathering NNAL data 42 . The main independent variable was the parents' smoking status, classified as "smoker" if any one parent replied "yes" or as "non-smoker" if both parents replied "no" to the question "Do you currently identify yourself as a smoker?" Independent variables that were considered to be potential confounding variables included sociodemographic, economic, and health-related characteristics, as well as the survey year. Sociodemographic characteristics included children's sex, age, parents' age, education level, type of housing, and region. Economic characteristics included the parental household income and ownership of private health insurance. Health-related characteristics included parents' drinking status and children's exposure to SHS at home and in public.
Statistical analysis. Univariate linear regression was used to assess the relationship between children's NNAL and cotinine concentrations and parents' smoking status; sociodemographic, economic, and healthrelated variables; and survey year. Prior to the multiple logistic regression analysis, we performed a log-transformation of the NNAL and cotinine values to ensure normality. Multiple regression analysis was performed while controlling for covariates to analyze the association between parental smoking status and log-transformed NNAL and cotinine concentrations in children. We performed subgroup analyses stratified by the parents' smoking status and multiple regression analysis to examine the associations of children's NNAL and cotinine concentrations according to the children's sex and BMI and the parental household income and education level. Furthermore, after adjusting for covariates, we classified the number of cigarettes smoked (0, < 10, 10-19, and ≥ 20) and smoking patterns (non-smoking parents, father only, mother only, both parents) and determined their associations with children's NNAL and cotinine concentrations using multiple regression analysis. All statistical analyses were performed using SAS software, version 9.4 (SAS Institute, Inc.). Statistical results were considered significant at a P-value of < 0.05.